NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/comp/unix/bsd/9059

Internet Message Format | 1992-11-18 | 1.9 KB

Path: sparky!uunet!sun-barr!cs.utexas.edu!zaphod.mps.ohio-state.edu!rpi!batcomputer!munnari.oz.au!comp.vuw.ac.nz!canterbury.ac.nz!otago.ac.nz!david
From: [e-mail] (David Harris)
Newsgroups: comp.unix.bsd
Subject: [386BSD] Warnings rebuilding libc.a
Message-ID: <[e-mail]>
Date: 19 Nov 92 13:48:01 +1300
Organization: University of Otago, Dunedin, New Zealand
Lines: 33

Maybe I'm just paranoid, but...

In the process of installing the crypt security support, I made the mistake
of rebuilding libc.a instead of simply replacing crypt_dummy.o with crypt.o
- for anyone out there contemplating this, it took nearly three hours and
2 crashes on a 486/50 to rebuild the libs - I don't recommend it unless you
have to do it.

However, in the process of rebuilding both the standard and profiled libs,
I noticed a number of warnings or perhaps errors from ar, of the general
form "cyclic data" - there were about 20 of them all at once, but they
scrolled past too fast for me to note down. As well, the resulting libc.a
was about 11KB larger than the bindist version.

Because of these warnings, I didn't want to risk installing the rebuilt
libc.a, so I simply patched the old one. Am I being paranoid? Are these
warnings in fact benign?

PS: Anyone know why all the ethernet cards are defaulted to IRQ2? On other
systems (NetWare and DOS) they warn you not to select IRQ2 for ethernet
cards in VGA systems because of possible conflicts with the VBL interrupt.
What's more, I don't think any of the supported cards uses IRQ2 as a
default out of the box (usually IRQ5 is safest, although most default to
IRQ3 which conflicts with COM2).

Cheers!

-- David --

=-=-=-=-=-=-=-=-=-=-=-=-=-=|  David Harris  |=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Consultant, Computing Services Centre,              phone: [phone]
University of Otago, Dunedin, New Zealand.          fax:   [phone]
   Internet email: [e-mail] (preferred), or [e-mail]